The
Linley Wire
Independent
Analysis of the Networking-Silicon Industry
Volume 5, Issue 10
May 19, 2005
|
 |
Editor: Linley
Gwennap
Contributors: Bob Wheeler, Jag
Bolaria
In
This Issue
Don't forget
to order the new edition of "A
Guide to Switch Fabrics" with the prepublication discount of
$300. This offer expires at the end of May.
Raza
Discloses Powerful Processor, EoS Chip
At Spring Processor Forum, stealth startup Raza Microelectronics
(RMI) announced its much-anticipated multicore processor, known
as the XLR. This powerful device contains eight MIPS64 processors
operating at speeds up to 1.5GHz. Although it was announced only
this week, the XLR has been sampling since December.
Like most network processors, the XLR uses multithreading to
improve the efficiency of its scalar CPUs while processing
many packets
in parallel. (David Hass, the XLR's architect, was formerly an
architect at Nexsi, a failed NPU startup also funded by Raza.)
Each XLR CPU switches among four threads, for a total of 32 threads
on the chip. The XLR is the only commercially available MIPS
processor to implement multithreading.
Along with its high-speed CPUs, the XLR contains 2MB of cache
memory and a crypto engine capable of 10Gbps of AES, 3DES,
SHA-1, or MD5
encryption. It supports 12.8GB/s of peak bandwidth to DDR2
SDRAM or RLDRAM. The processor connects to the rest of
the system through
two integrated 10GbE MACs or four integrated GbE MACs as well
as other high-speed interfaces. RMI quotes a list price of
$850 for
the high-end XLR but offers lower-cost versions with fewer
CPUs and other restricted capabilities.
The company positions the XLR for a wide range of control-plane
and data-plane applications. The chip is well suited for
high-end control-plane designs; compared with Broadcom's
BCM1480, the
XLR offers more CPUs, more clock speed, and multithreading
for a lower
price. We do not, however, expect the XLR to displace traditional
NPUs from Layer 3/4 routing applications, due to the XLR's
greater cost and power dissipation and incomplete software
solution.
Instead, we see the XLR competing with Cavium's Octeon in
data-plane applications
that require greater per-packet processing, such as security
(intrusion prevention/antivirus), IP storage, and web switching.
Octeon lacks
the XLR's multithreading capability but offers a hard-wired
DFA engine, useful in many security applications.
At the same time, RMI also introduced its Orion chip, designed
to provide packet service over the Sonet/SDH network. The
Orion family includes products for OC-3, OC-12, or OC-48
Sonet rates
and 8 FE ports, 2 GbE ports, or 3 T3/E3 packet ports. The
Orion architecture consists of a Sonet/SDH mapper and an
Ethernet
interface. The mapper includes a Sonet/SDH framer, pointer
processor, VCAT
(virtual concatenation), LCAS, and GFP framer. The Ethernet
interface includes MACs, MAC forwarding, and a traffic
manager for 2K flows.
The chip may be used for Ethernet over Sonet (EoS) or to
switch Ethernet traffic.
Although
vendors such as PMC-Sierra and TranSwitch have been shipping
data-mapping products for similar applications,
RMI offers a highly
integrated chip with sophisticated traffic-management
capabilities that include dual-bucket policing, random
early discard,
and shaping. As a late entrant, RMI's challenge is to
use Orion's
unique features
to establish a position in this nascent market. —LG/JB
Complete
coverage of Cavium's Octeon and similar processors appears in
our report A Guide to High-Speed Embedded Processors.
Cortina
Rolls Out RPR Products
Cortina Systems has announced availability of its second generation
of RPR (Resilient Packet Ring) products. Cortina, a leading supplier
of RPR silicon, already has design wins at Cisco and several major
OEMs, including Extreme, Huawei, and Alcatel. At Cisco, the startup's
technology is deployed in shared port adapters or I-Flex, which
enable a single line cards for different networking protocols (ATM,
Ethernet, Sonet). Madrid, the company's newest device, is a dual
RPR MAC for applications such as ADMs, MSPPs, switches, and routers.
Madrid enables TDM networks to be used for shared packet services
with quality-of-service guarantees. It can also be used for failure
protection and guaranteed service for packet networks in metro
and enterprise networks.
Madrid is a 10Gbps RPR device that terminates one RPR ring
with two RPR channels, which can occupy any slice of the
10bps bandwidth.
The device provides either a SPI-3 or a SPI-4.2 interface to
enable 10Gbps and 2.5Gbps ADM/MSPP line cards. In these
applications,
Madrid would connect to an ASIC and Ethernet devices towards
the line side and to a VCAT (virtual concatenation) framer
towards
the system side. For legacy applications, Madrid supports SRP—the
precursor to RPR developed by Cisco.
Madrid
also integrates two GbE PHYs using TBI, which can connect to
an external serdes device. The Ethernet interface allows
Madrid to be deployed in applications that aggregate Gigabit
Ethernet
ports. Madrid uses SRAM to support enhanced bridging
for up to 128 MAC addresses. This feature is useful for
bridging
between LANs without flooding MAC addresses. The device includes
a mode
to bypass the RPR processors and MACs for up to 64 channels.
As
some established vendors have cut back on Sonet and RPR programs,
Cortina has continue to broaden its product lines.
In 2004, it
acquired Azanda for SAR and traffic-management capability
and now claims design wins in more than 45 line cards.
With limited
competition
and few new entrants, Cortina is well positioned to extend
its business by reusing a combination of its RPR, Sonet,
ATM, and
Ethernet technologies. —JB
Additional
coverage of Cortina and other EoS devices appears in our report,
A Guide to Next-Generation Sonet Silicon.
AMCC Offers Secure PowerPC Chips
This week, AMCC announced two new PowerPC processors, the 440GRx
and 440EPx. These processors are AMCC's first to integrate an encryption
function, achieving 500Mbps throughput on IPSec using 200-byte
packets. Using an encryption engine licensed from SafeNet, the
new chips can encrypt and decrypt IPSec packets with no CPU load,
freeing cycles for the main application.
The
new chips, due to sample in 4Q05, use the same PowerPC 440 CPU
as the earlier 440GR and 440EP but boost the clock
speed as
high as 667MHz. They also add a Gigabit Ethernet port to the
other peripherals already included in earlier family members
and support
the new DDR2 SDRAM standard. With pricing as low as $20 for the
440EPx, the new processors target multifunction printers, wireless
access points, residential gateways, and other SOHO equipment.
—LG
Coverage
of competing devices appears in our report A Guide
to Communications Processors.
The
Linley Group Seminar on Access System Design
Join us on July 21 when The Linley Group hosts a free one-day
seminar on access system design at the Marriott Hotel in Santa
Clara. This event will be similar to our successful seminar
on fabrics and high-speed interconnects, which was attended
by more than 100 people.
The access seminar will consist of four sessions. The first
session, featuring The Linley Group analysts, will be an overview
of challenges in packet processing for access-equipment designers
and the various technologies available for solving these issues.
The second
session will describe and compare access-focused CPUs such
Freescale’s PowerQUICC and AMCC’s PowerPC
processors. The third session will cover access-focused NPUs
such as Wintegra’s WinPath and AMCC’s nP3xxx. Each
of these two sessions will include technical presentations
from leading vendors as well as a moderated panel discussion.
The seminar will conclude with a panel discussing the longer-term
trend of converging CPU and NPU functions into a single chip.
The event is designed to provide in-depth coverage of the
most recent advances in access-system design presented
by industry
experts such as The Linley Group analysts as well as architects
and chip designers from leading vendors. The event includes
lunch and a reception to provide opportunities for you to
network with these industry leaders, analysts, and your
colleagues.
This seminar is free to qualified individuals who register
by July 15. The seminar is targeted at OEMs, VARs, press,
and the financial community.
Click
here for further details and registration
information.
Event sponsored by Freescale, Wintegra, AMCC, and The
Linley Group.
Call for Papers: Network Systems Design Conference
We are now accepting proposals for presentations at the Network
Systems Design Conference, which will be held in San Jose,
California on October 18-20, 2005. NSDC will highlight the
newest technologies and chips for networking applications.
The only conference totally dedicated to network systems design,
sessions will include technical disclosures of new networking
chips as well as in-depth discussions on how to improve the
design of networking equipment.
NSDC focuses on technical presentations and on new disclosures.
Proposals will be judged by the amount of new technical content
(disclosed at the show or within the previous couple of months)
and its significance to our audience.
We are seeking proposals discussing network processors
(i.e., chips that process packets) in all performance
ranges and
feature sets, NPU software and tools, communications processors,
wireless
LAN chip sets, storage processors, control-plane processors,
security processors, classification coprocessors, search
engines, and switch fabrics as well as interface, controller,
and switch
chips for Ethernet, Sonet, ATM, and Fibre Channel.
Proposals may discuss standard products or in-house ASICs
as well as techniques or technologies for using these
products, however, preference will be given to commercial
or commercializable
technologies. We do not require that chips have reached
first silicon, however, all presentations must include
a significant
amount of specific technical details about the product
as well
as an estimated availability date.
Proposals should include the name of the device or system
to be discussed, a general description of the device,
a list of
specific features that will be discussed in the presentation,
the name and title of the presenter, and contact information
for the person who should be contacted regarding the
proposal. We strongly prefer presenters from engineering
(e.g., chief
architect, CTO) rather than marketing. In addition,
the proposal should detail what information will be
disclosed
at NSDC
2005; for information that will be announced before
NSDC 2005, please
indicate when it is scheduled to be announced.
The
deadline for submitting proposals is June 24, 2005. For more
information, or to submit a proposal, access
the NSDC
web site or e-mail
your proposal to program@networksystemsdesign.com.
The Linley Group Seeks New Analyst
The Linley Group is looking for a senior person to add to
its analyst staff. The ideal candidate has both engineering
design and marketing experience, able to discuss both the technical
and business aspects of a product. Strong communications skills,
both written and presentation, are required. Background in
processors and/or networking is preferred.
This position offers a unique vantage point on new technology
developments across the networking-silicon industry. The
Linley Group analysts are well known and respected.
We offer the flexibility
of a small company with health and retirement benefits. Compensation
is commensurate with experience and ability. For more information,
submit your resume to jobs@linleygroup.com.
Principals only please. No agency candidates will be considered.
No phone calls.
To
receive The Linley Wire via e-mail, click
here
About
The Linley Wire
© 2002-2005 The Linley Group
|